Explosive expansion of data consumption has led to expanded demand to process and analyze data utilization. Market and business requirements such as service improvements, capacity assessments, and customer demand continually raise data consumption metrics gathering and analysis demands. Online services provide a unique opportunity in allowing direct access to utilization metrics. Gathered metrics are available to multitude of analysis schemes for system study and evaluation. In return, analysis schemes impact various business needs such as targeted improvements, return on investments, and demand predictions.
In modern systems, quality of service (QoS) and usage reporting are two service areas, where online services excel at analyzing and providing consumption metrics. Components of online services have different demands for quality of service and usage reporting. System architectures (e.g. message based, session based, request based, etc.), usage patterns (user-initiated, scheduled processes, asynchronous processes), and application layers (externally visible, internal service) may set difficult demands on QoS and usage reporting systems.
Conventional distributed data processing systems provide different approaches to performing analysis and aggregation of large quantities of usage data. However, such systems are distributed batch processing systems and are typically not online systems suitable for reporting.